Hindi_base_wav2vec2 Open-Source Hindi Speech Recognition Model - Free Deployment for Accurate Speech Recognition

Hindi Base Wav2vec2

Developed by Harveenchadha

This is a Hindi automatic speech recognition (ASR) model based on the wav2vec2 architecture, trained on datasets like Common Voice.

Downloads 49

Release Time : 3/2/2022

Model Overview

This model is specifically designed for Hindi automatic speech recognition tasks, capable of converting Hindi speech into text.

Multi-dataset training

The model was trained on multiple versions of Hindi datasets including Common Voice 7.0 and 8.0.

Good performance

Achieved a WER of 19.47-22.62% and CER of 7.42-9.47% on the Common Voice test set.

Open-source license

Released under the Apache-2.0 license, allowing both commercial and research use.

Hindi speech recognition

Speech-to-text

Speech transcription

Hindi speech to text

Convert Hindi speech content into editable text

Approximately 80% accuracy (WER around 20%)

Voice assistants

Hindi voice command recognition

Provide speech recognition capabilities for Hindi voice assistants

Property	Details
Model Type	Hindi Large
Tags	automatic - speech - recognition, hf - asr - leaderboard, hi, model_for_talk, mozilla - foundation/common_voice_7_0, robust - speech - event
Datasets	Harveenchadha/indic - voice

Task	Dataset	Test WER	Test CER
Automatic Speech Recognition	Common Voice (hi)	22.62	7.42
Automatic Speech Recognition	Common Voice - 7.0 (hi)	19.47	8.05
Automatic Speech Recognition	Common Voice - 8.0 (hi)	20.87	9.47

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base